We can find more papers on the web than we used to. Are we reading them?

 

STUDENT: What did you do before the web to find papers?

BILL: We went to the library and copied papers to read later.

STUDENT: Did you read them later?

BILL: Well,  uh,hmm,  ...

BILL to a professor in his 80's: What did you do before copy machines?

PROF: We went to the library, took the journal off of the shelf, and read the article. We may also have taken notes on paper on the article. 

Back to 2025: I do the following sequence of events with a branch point at the end.

1) Get interested in some problem. 

I give an example:

 Rado's Theorem for equations  with a constant. Short summary:

Rado's Theorem: Let \(a_1,\ldots,a_n\) be integers. The following are equivalent

--For all finite colorings of \(N^{\ge 1}\) there is a monochromatic solution to \(\sum_{i=1}^n a_ix_i= 0\)

--Some subset of \(a_1,\ldots,a_n\) sums to 0.  

My interest: What if we look at equations which replace 0 by some other constant? AND what if you just want to look at (say) 2-colorings? 

2)  Assemble a website of papers on the topic.

I assembled a website of papers on the Rado question. I did that. The website is   here. 

3)  I then read the papers and understand them. Or not. 

I intend to do the following: 

a) Read the paper with pen and pad and take notes, draw diagrams, do examples. I may create some  handwritten notes. The benefit of the notes is from making them. The actual notes are terrible. 

b) Type in the notes and while doing that polish them. I might work with a student on this.

c) If needed make slides out of the proof for presentation.

I sometimes go straight from the paper to the slides. I've stopped doing that since it is valuable to first go through the Understanding the proof  phase before going to the Best way to present the proof phase.

----------------------------------------

Step 3 can be a problem. I have many website of papers that I never got around to reading. The key for me is to strike while the iron is hot that is, SOON after assembling the papers READ them.

 For Rado's Theorem with constants I have not gotten around to reading any of the papers yet. 

There is some benefit to having to read a paper in the library since then you actually read it. I am NOT saying  things were better when I was a kid. The lesson is more of how to combine current technology with the helpful constraints of a prior era. 

Another approach: go to the library and look through journals to find something interesting. I've done this a few times with success. I blogged about one of them  here